Dear Mihir,
thank you for your effort, see my comments inline:
On Sun, Apr 28, 2013 at 6:17 PM, Mihir Rege <[email protected]> wrote:
> Hi everybody,
>
> I have been working on the idea for creating an interface for tagged
> corpora since the past few days. I have finished the coding challenge, have
> gone through the documentation pdf and wikis and am currently working on
> designing the interface. I am posting the mockups for the interface so
> that I can get your opinion and improve on them.
>
Great!
> There are currently three major interfaces:
>
> a) Manual disambiguator
>
> b) .prob evaluator
>
> c) .tsx file editor
>
> I have put up all the mockups together at http://imgur.com/a/4uk4q#r5Ur8jT and
> have also put links in separate sections.
>
>
>
Let's take a look, I like the idea of having access to all possible tasks
that apply to work with taggers in an interface, it is difficult to
organise because the process can be non-linear but still, a good idea.
> *a) Manual disambiguator*
>
> Mockup: http://i.imgur.com/r5Ur8jT.png
>
> Functions:
>
> - Jump to next ambiguous lexical unit or adjacent lexical-unit using
> the keyboard or mouse.
> - A quick-view bound to a key, to hide the tags and show the raw text
> - If the .tsx file is provided, information like the coarse tags,
> forbid, enforce rules applicable can also be displayed.
> - Show statistics of disambiguation
> - Compile and apply constraint grammar rules to the buffer
> - List the applied constraint grammar rules
> - Train and test the tagger (a prompt will ask the part of the corpus
> to be used as testing data).
> - Train the tagger and export the .prob file
> - Save progress ( this will save the corpus and also create a project
> description file which will keep track of the morphological analyser, .tsx
> files used, so that it is easier to resume tagging)
> - The interface will be keyboard centric, though it will be equally
> functional with a mouse.
> - Default keymaps will be provided and the bindings can be changed to
> suit the user
>
> For example
>
> [P] - <previous-ambiguous>
>
> [N] - <next-ambiguous>
>
> [F] - <forward-word>
>
> [B] - <back-word>
>
> [1], [2],[3],[4] for choosing the correct lexical form.
>
Cool functions, all very useful for the end-user of this tool. I think that
one more function (if possible) would be really useful: select one or more
words and see the output of the morphological analyser for them and be able
to change the tagged corpus according to the morphological output. This
will help a lot to re-tag the corpus to make it consistent with changes in
dictionaries.
*Evaluating the tagger*
>
> Functions
>
>
> - The trained tagger can be evaluated immediately by having an option
> of setting aside x% of the corpus as testing data.
> - Else, it can be evaluated using the .prob evaluator using an
> unrelated corpus.
>
> That's fine. Many times what we want to do is to compare two .probs to see
if the "improved" one is doing better than the "old" one. We should maybe
provide some way to make it easy: test with two .probs and show differences
to the user.
*Loading the corpus*
>
> The available options are:
>
> 1.
>
> Load a raw-text file, morphological analyser and .tsx file (optional)
> 2.
>
> Continue on an existing project
> 3.
>
> Pull a wiki-dump and use it as the corpus [
> http://i.imgur.com/F9OXMs4.png ]
>
>
>
> I think we should add something about the language. The user should be
able to define the language of the raw-text file or of the wikipedia dump,
then have a list of available pairs or languages (if any) to create the
first tagged but non-disambiguated version of the corpus.
>
> 1.
>
>
> *b) .prob evaluator*
>
> Mockup: http://i.imgur.com/fIo6rV9.png
>
> Functions
>
> - Input the .prob file , the manually disambiguated corpus along with
> morphologically analysed corpus or the morphological analyser for the
> language.
> - Evaluate the .prob file and display statistics about tagger accuracy
> - Generate a log file, which will basically be the diff between the
> provided tagged corpus and the corpus disambiguated by the tagger, making
> it easier to frame new sentences to add to the corpus, so as to give more
> context to the tagger
>
>
Great, see my comment about comparing two .probs above.
>
>
> *c) .tsx file editor*
>
> Mockup:
>
> TSX Viewer
>
> http://i.imgur.com/pVdsIem.png
>
> Templates
>
> http://i.imgur.com/hFGIQHR.png
>
> Functions:
>
> -
>
> Add new tags
> - categories
> - multi-categories
> - forbid
> - enforce
> - prefer
> - Templates for adding new tags
> - Change the order of the tags (as more specific categories must be
> defined before more general ones) within the same parent tag. The nodes in
> the xml viewer can also be made draggable within the same parent node to
> make it easier to change the order
> - Search within tags for faster navigation.
> - Validate the tagger definition
> - Editor features like syntax highlighting , auto-indentation and tag
> completion for manual editing in the Node Contents textview for complex
> in-place editing.
>
>
> Pretty good. A more human-oriented representation of the information
(symbols meaning) could be also a good thing to have in this interface. And
if we can produce examples that represent what a coarse tag groups or what
a forbid, enforce or prefer rule is doing, that would be brilliant as well.
> Looking forward to hear from you. :)
>
>
>
Sorry again with the delay and don't forget about adding your proposal.
> Regards,
>
Best,
Gema Ramírez.
> Mihir Rege,
>
> Second Year Undergraduate,
>
> Department of Computer Science and Engineering,
>
> IIT Kharagpur.
>
>
>
> ------------------------------------------------------------------------------
> Try New Relic Now & We'll Send You this Cool Shirt
> New Relic is the only SaaS-based application performance monitoring service
> that delivers powerful full stack analytics. Optimize and monitor your
> browser, app, & servers with just a few lines of code. Try New Relic
> and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
> _______________________________________________
> Apertium-stuff mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/apertium-stuff
>
>
--
Gema Ramírez
---------------------
Prompsit LE
Traduce, extrae, analiza: http://aplica.prompsit.com
------------------------------------------------------------------------------
Introducing AppDynamics Lite, a free troubleshooting tool for Java/.NET
Get 100% visibility into your production application - at no cost.
Code-level diagnostics for performance bottlenecks with <2% overhead
Download for free and get started troubleshooting in minutes.
http://p.sf.net/sfu/appdyn_d2d_ap1
_______________________________________________
Apertium-stuff mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/apertium-stuff
